On 5/16/2017 1:45 PM, Booth Martin wrote:
This leads to another point that has bothered me. How is the inventory
of Service Program functions documented?
Many years ago, I adopted a naming convention I saw David Morris use.
I keep my prototypes in a separate source file called QPROTOSRC. So
it's easy to find all of the prototypes I ever wrote; they are in one
source file, in a relatively small number of members.
In addition to that, I wrote a keyword documentation thing that scans
the text in these members and generates a simple list of keywords and
the source member they are in. It's easy to search that index for words
that strike my fancy. This one's for when I have a vague recollection
of what I wrote but can't exactly remember.
In addition to that, I've been keeping a wiki on my own system (I run
both TWiki and MediaWiki on IBM i) where I can document this stuff in a
more formal manner.
Finally, I wrote another tool that uses an API to extract all the
sub-procedure names from all the service programs on the system and put
that into an index that I can easily search to find where a given thing
is encapsulated. This one's for when I'm looking at a mainline program,
it's using ibeamWeight(), and I have no idea which service program I put
it into.
midrange.com
